International Journal of Engineering and Applied Sciences (IJEAS) ISSN: 2394-3661, Volume-2, Issue-2, February 2015 An Empirical Analysis of XML parsing using various operating systems. Amitesh Saxena, Dr. Snehlata Kothari used everywhere in software. An XML Parser is a parser that Abstract— As the use of internet technologies are widely is designed to read XML and create a way for programs to use increasing, the XML markup language attains a remarkable XML. There are different types, and each has its advantages. importance due to its language neutrality and independency in Unless a program simply and blindly copies the whole using data exchange and data transfer through web XML file as a unit, every program must implement or call on environment mechanism. For improving the processing an XML parser. performance of XML parser, it is necessary to find out a mechanism, in which we get minimum processing time while B. DOM (Document Object Model) parsing of XML documents. In this paper, XML documents are being experimentally It supports navigating and modifying XML documents tested using various operating systems to determine, whether an - Hierarchical tree representation of document operating system effect the processing time of XML parsing. - Tree follows standard API - Creating tree is vendor specific Index Terms— XML Parser, DOM Parser, Operating DOM is a language-neutral specification system. - Binding exist for Java, C++, CORBA, JavaScript, C# - can switch to other language. I. INTRODUCTION The Document Object Model (DOM) is In Present world, mega information are sharing and an interface-oriented application programming interface that transmitting, In this XML plays a very significant role as a allows for navigation of the entire document as if it were a tree worldwide design for data interchange. It allows users to of node objects representing the document's contents. A share XML documents. XML is capable to the mining of data DOM document can be created by a parser, or can be from an XML document without any facts or knowledge generated manually by users (with limitations). about the contents of that. XML documents need to be conformant with XML specifications for achieving this transparency. By using an XML parser, this specification III. LITERATURE REVIEW conformance can be checked. The parser makes the data easy to get and also ensures the validity of that. Review of work already done on the subject A study of various XML parsers could be very positive in In today’s, a large number of XML parsers, coded in a verity determining the strengths and weaknesses of various XML of languages, May be these parser not give the similar parsers in respect to the different features of XML. Parsers performance in terms of parsing speeds, accuracies, and can be compared to one another by checking their storage requirements. This is lastly proofed that, for execution conformance with the XML recommendations [1] given by time savings, accurate parsing, and storage requirements, a the World Wide Web Consortium. The Organization for parser must be selected to fit those specific requirements. Advancement of Structured Information Systems (OASIS) [2] is a NGO that has collected a number of test cases from In this study, XML document will be tested with three DOM various sources and built a Conformance Test Suite for XML API, i.e. PHP,JAVA and Microsoft on various Operating with approx. 2000 test cases (as of November 6, 2001). Systems like WIN7, WIN8, UBUNTU, Red Hat, etc. The main objective of this study is to check whether the operating Anez [3]. 1999, conducted a study to determine the suitability system affects the parsing speed and if yes then the best of XML and Java for the representation and manipulation of combination of parser and operating system will be also Transport and Land Use (TLU) modeling information as used finding out. in urban and regional planning. This study evaluated seven different XML parsers with respect to conformance with II. TECHNOLOGY OVERVIEW XML specifications, speed and memory usage. In this study, the test suite developed by James Clark (currently part of In this section we describe about different XML Parsers. OASIS Test Suite) was used for checking the conformance of A. XML Parser different parsers. Speed and memory usage tests were A parser is a piece of program that takes a physical performed using two large XML files (0.8 and 1.2 MB, representation of some data and converts it into an respectively). Each of these two files contained several in-memory form for the program as a whole to use. Parsers are thousand XML elements nested in a four-level deep hierarchy, and all of the elements had one or more attributes. The study concludes by giving rankings to the different Amitesh Saxena, Research Scholar of Pacific University, Udaipur parsers as shown in Table 1. Dr. Snehlata Kothari, Pacific University, Udaipur. 37 www.ijeas.org An Empirical Analysis of XML parsing using various operating systems. Table 1: Rankings for different parsers by Anez [3] Takase et al. [13]. 2005, explores a different way to improve Parser Rank parser performance. It memorizes parsed XML documents as IBM XML4J (XML for java) v Outstanding byte sequences and reuses previous parsing results when the 1.1.4 byte sequence of a new XML document partially matches the memorized sequences. James Clark’s XP v0.4 Good Sosnoski [14]. 2005, carried out a test on DOM based parsers using XMLBench. He tested on the execution speed and Microsoft XML (MSXML) v Good memory usage for a set of XML documents ranging from 1.9 small-scale to large-scale file sizes. The test result shows that Microstar Aelfred v1.1 Good Xerces outperforms among the others. Besides, Xerces parser is also voted as the best XML parser of the year by Sun XML (under construction) Acceptable XML-Journal/Web Services Journal Readers' Choice Awards [15-2004].Since Xerces and MSXML outperform the rest of the parsers in most cases, we have decided to concentrate Loria sxp v 0.72 Acceptable benchmarking our proposed parser, xParser against these two parsers. Data Channel XML parser Poor Perksins et al. [16]. 2005, where authors use a small (less than 1 KB) XML representing a typical purchase order This study also compares the relative performances of the structure to test transcoding impact and object creation of parsers with respect to conformance with XML DOM, SAX and JAX-RPC. The authors also explore the recommendations, speed, and memory usage. It concludes navigation costs of each API and compare the results with a that different parsers excel under different requirements. specific XPath parser. However, this study does not provide quantitative data about A study towards different XML parsers is beneficial when the conclusions. comes to determine the strength and weaknesses of the Claben [4]. 1999, considered the following parsers: IBM products. Various studies have been conducted which XML4J, Apache Xerces, Sun Project X, Microsoft MSXML, compare on conformance to standards, speed, memory usage Oracle XML parser for Java, and James Clark XP. He used and so on. the following features for comparing the different parsers: Lam T.C., Ding J.J. and Liu J.C. [20]. 2008, concluded that Well-formedness, Validity, XML Schema, Namespaces, the process of handling XML documents was described in XSL-T, SAX levels 1, 2, and DOM levels 1, 2. The focus of four phases: Parsing, that is considered a critical step in this study was to determine the set of features supported by performance, Access, Modification and Serialization (figure each of the parsers using the OASIS Conformance Test Suite 1), whose performance is directly affected by the parsing (1000 test cases1). The study concludes that Sun's parser and models. James Clark XP are the best parsers supporting the XML standard, but it does not give any quantitative figures of how good or bad each parser is with respect to a specific feature. Cooper [5], 1999 studied how parsing speeds vary with the programming language used for developing the parser. In this study, two java parsers, two C parsers, one perl and one Figure 1. Example of a XML memory tree representation python parser were used. Five XML documents with sizes ranging from 160 K to 5.0 MB were used in this study. This As the most critical factor of performance, parsing is study concluded that C parsers are always faster compared to characterized by the conversion of characters, mainly related java, perl or python parsers. to the conversion of characters into a format that a Mohseni [6]. 2001 performance test indicates that MSXML programming language understands, lexical analysis which is rivals other parser having the shortest loading time. the process that identifies XML elements, e.g. start node, end Noga, M., Schott, S., L¨owe, W. [8]. 2002, Therefore efforts node or characters, applying regular expressions defined by have been made to improve DOM parser performance by World Wide Web Consortium (W3C) . The last step of the exploiting lazy XML parsing. The key idea is to avoid loading 1 parsing phase is the syntactic analysis of the document, where unnecessary portion of the XML document into the DOM it is checked if the document complies with the rules of tree. It consists of two stages. The pre-parsing stage builds a construction of an XML document. Finally, the API virtual DOM tree and the progressive parsing stage expands implements access and modification operations on the data the virtual tree with concrete contents when they are needed resulted from the parsing process. by the application. They analyzed parsing models, data representations and their Oren [9]. 2003, proposes Piccolo XML parser presenting a impact on XML processing.